Page 1


Gemini Legacy Image / NGC 6946 / T. Rector, University of Alaska Anchorage

Gemini Legacy Image / HGC 92 / T. Rector, University of Alaska Anchorage



Gemini Legacy Image / NGC 1532/31 / T. Rector, University of Alaska Anchorage

Gemini Observatory


A Look Inside 4 The Next Decade at Gemini Observatory A look at the future of Gemini and the process that shaped our vision.

6 Gemini Campaign Science


f you are a regular reader of the Gemini Observatory Newsletter this issue is probably a bit of a surprise. In an effort to better serve our user community the

Newsletter has become

GeminiFocus – with a new

name, look and approach that are a radical departure from the past. We hope you will find the full-color design created by Gemini Graphic Artist Kirk Pu‘uohau-Pummill

As new instruments become reality,

much cleaner and easier to read. The editorial content

Gemini plans large observing campaigns to

has also changed significantly, with greater emphasis

answer compelling astronomy questions.

11 Beta Pictoris with T-ReCS Asymmetries and dust particle emissions

point to possible recent collisions around this nearby star.

on science content and feature articles that are more appropriate to our role as a fully operational observatory. The future of the Gemini Observatory is unfolding and

GeminiFocus reflects our evolution.

14 Accretion Signatures in Massive Stars Gemini South and NOAO’s Phoenix spectrometer probe deep into massive GMOS-N / I. Braul, Southlands Elementary, Vancouver, Canada

young stellar object environments.

16 Dust Emission from Massive-Star Supernova Gemini North observations single out a supernova in NGC 6946 for further study.

23 Key Problem in Stellar Evolution Tackled Data from three telescopes on Mauna Kea probe mass loss in intermediate-mass stars.

27 Did The Most Massive Galaxies Form First? Astronomers use Gemini to visit the “redshift desert” and search for distant, massive galaxies.

30 Recent Science Highlights

Recent science results highlight galactic black holes, a planetary nebula, and a brown dwarf.

Managing Editor


Peter Michaud

Science Editor | Jean-René Roy Associate Editor | Carolyn Collins Petersen Associate Science Editor | Rachel Johnson Designer | Kirk Pu‘uohau-Pummill



by Matt Mountain & Jean-René Roy

Gemini Observatory: the Next Decade A

t a retreat in Oxford, UK in September

The process was initiated in November 2002 when

2004, the Gemini Board concluded

the Gemini Board established a Strategic Planning

a three-year, intensive and rigorous

Group with the purpose of creating a vision for

examination of the Observatory’s future. The

Gemini’s future. Figure on fold-out highlights the

Board, representing the scientists and agencies

key events and results that followed.

of the Gemini partnership, issued the following statement: We see the Observatory establishing a leadership role in a global effort to define, address, and solve compelling scientific questions. The answers to these questions will have a fundamental impact on our view of the Universe and our place in it. Gemini, by exploiting its unique strengths and capabilities, will be a keystone in that global effort. Among

Astronomers from across the Partnership at the Gemini Science Conference, June 2004

our strengths are the breadth of the partnership, the diversity

It is not just the activities described on the fold-out

and depth of our communities and staffs, our connections

(often called the “Aspen Process”) that are changing

with other institutions sharing common scientific aspirations,

the way Gemini will operate. As a result of this

and the energy and vision of our Observatory.

continuous consultation and review process, the Gemini Board (as reported in the previous Gemini

The Board also specified a number of changes

Newsletter, December 2004, issue 29) has asked the

in observatory operations while endorsing the

observatory to initiate a number of key changes in

Gemini community’s scientific roadmap for the

the way we operate for 2006 and beyond:

partnership, encapsulated in the Aspen Report, “Scientific Horizons at the Gemini Observatory: Exploring

It’s clear from the increased proposal pressure

a Universe of Matter, Energy and Life” available at

through our National Time Allocation

Committees that the original plan of offering 50% queue and 50% classical has not given the


As we approach 2006 and conclude the last year

community or the observatory enough

of our current operating plan, a new and vigorous

flexibility in optimizing the scientific output

program for 2006-2010 and beyond has been

of the two Gemini telescopes. We are now

established. It is important to understand how we,

working to support both telescopes for 100%

and our communities, arrived at this pivotal point.

of nights, giving our communities opportunities

Gemini Observatory

June2005 to experiment in new and innovative ways with

Subaru Observatories as described on page 40 of

the observing process. These include “real-time”

this Newsletter.

queue observing through “remote

eavesdropping,” arbitrary-length classical or

The impact of the Aspen Process and the

Committees and reports that have contributed to the Gemini vision

semi-classical “assisted” runs, and fully

transformation of our operating model and partner

for the next decade:

supported Target of Opportunity observations

national offices has been a three-year effort

for research in areas such as supernovae and

involving all Gemini partners. In particular, the

gamma ray bursters.

Gemini Board and observatory have assembled

As instruments and their data sets become more

input from its Science Committee, Science

complex, our users are looking for more

Working Groups, the National Science Advisory

complete data sets with the telescope and

Committees, our AURA Oversight Committee, the

instrument signatures removed (as is now the

Aspen and Gemini Science Workshops, the Visiting

“norm” for facilities such as HST, Chandra,

Committee, and Mid-Term Management Review

Spitzer and the VLT). Thorough and automated

Committee; collectively representing a broad and

reduction of complex data sets immediately

continuous consultation with our communities.

useable for research has now become a practical,

By working together, we have constructed a

and as it turns out, desirable goal for us to

step-by-step consensus on where Gemini and

better enable the scientific aspirations of our

its users should head in the next decade. Due

user community.

to this unprecedented collaboration between the

Gemini operates a unique user support model

observatory, its community and agencies, we now

with local, national support expertise resident

have an exciting scientific future ahead of us!

Menbership Gemini Board

Gemini Science Committee Operations Science Working Group Adaptive Optics Science Working Group Wide Field Optical Spectroscopy Science Working Group National Gemini Offices AURA Oversight Committee for Gemini Gemini Visiting Committee & Mid-Term Management Review

Aspen Workshop Vancouver Gemini Science Conference

(NGOs). To make this network more effective,

Preliminary Report of the Strategic Planning Group, May 2003 Message from the Chair of the Gemini Board, December 2004 Resolutions of the GSC

Report of the Gemini Visiting Committee and Midterm Management Review Committee • Response of the Gemini Board Scientific Horizons of the Gemini Observatory Program •

All of these documents can be accessed at:

Gemini and its partner NGOs are working together to produce a more seamless, coordinated and accountable user support “system of systems.” The resources required to stay competitive in every aspect of optical/infrared ground-based 8- to 10-meter-class astronomy is today a considerable challenge for any one major observatory. However, if collectively we can give our respective communities access to a “system” of 8- to 10-meter telescopes through a greater degree of coordination and collaboration this will allow our collective communities to maintain access to a broader range of the

Representatives of the Gemini Community at the Aspen Workshop, June 2003

competitive capabilities in the coming decade. The first tentative steps in this “experiment” are now being taken by the Gemini, W.M. Keck and


in each country in “National Gemini Offices”

Relevent Reports /Resolutions


Gaining a Consensus from Across the Gemini Partnership

by Joe Jensen

May 2005 - Gemini Board endorses an Aspen package that would allow the realization of most of the Aspen science objectives over the next six to eight years. March 2005 - Multi-institutional teams from across the Gemini partnership present a comprehensive proposal for instruments and capabilities that would execute core Aspen science objectives. October 2004 - Gemini Board holds a retreat to “seek to agree on a shared vision of Gemini’s evolution and for provision of necessary resources within agreed limits.” Result: November 2004 Gemini Board establishes guidelines for 2006-2010 and begins work with funding agencies to identify resources that will realize the Aspen aspirations. June 2004 - Over 100 astronomers from all Gemini partners assemble in Vancouver, BC for a three-day science conference. Result: Almost 100 papers on Gemini results presented and a consensus emerges that the observatory should move to a higher fraction of queue-based observing.

Gemini Campaign Science

May 2004 - Observatory selects six partnership-wide teams to prepare Aspen design and feasibility studies. March 2004 - Combined NSF Visiting Committee/Mid-Term Management Review undertake a thorough examination of Gemini operations and development plans. Result: Recommendation to NSF and Gemini Board to invite AURA to propose an Operating Plan for 2006-2010 that would enable science capabilities identified in the Aspen process.

Astronomers (above) from around the globe converged in Aspen to think of ways to answer the most important questions in astronomy.


n June 2003, representatives of the Gemini

is the High-Resolution Near-Infrared Spectrograph

partners met in Aspen, Colorado to discuss

(HRNIRS) designed for precision radial velocity

the future direction of the observatory. The

studies to search for terrestrial planets around low-

participants in the Aspen meeting articulated

mass stars. HRNIRS would also have a second

science goals that included answering the deepest

multi-conjugate adaptive optics (MCAO)-fed

and most exciting questions we can currently

multi-object mode for high spectral and spatial

pose about the nature of the universe and

resolution studies of multiple objects in star

June 2003 - November 2003 - Gemini Science Committee works with the chairs of the Aspen panels to refine key capabilities required to address the Aspen questions. Result: Gemini Board releases Aspen Road Map: Scientific Horizons at the Gemini Observatory: Exploring a Universe of Matter, Energy and Life, (Table on previous page for URL access).

the characteristics that allow life to exist. The

forming regions. These two instruments would

workshop participants didn’t stop after posing those

directly address questions about how stars and

questions, but imagined the kinds of observations

planets form, and how common the conditions for

and types of instruments that could be used on the

life might be in the universe. The third instrument

Gemini telescopes to furnish the answers.

recommended by the GSC was the Wide-Field

June 2003 - Gemini and partner countries hold “Aspen Workshop.” Result: 93 astronomers participate in a two-day workshop in Aspen, Colorado. Toplevel science questions on the universe of matter, energy and life are identified that Gemini could answer in unique ways.

The concepts for future Gemini instruments

would measure 4,500 spectra simultaneously over

developed in Aspen were brought to the Gemini

a 1.5-degree field of view. This revolutionary new

Science Committee (GSC) for further refinement

capability would allow Gemini astronomers to

and prioritization. The GSC endorsed the Aspen

attack questions about the nature of dark energy

science goals, and suggested that a small number

and dark matter, as well as the origin and eventual

of instruments addressing the majority of the

fate of the universe. WFMOS would also survey

Aspen science questions be built. One of the

millions of stars in our galaxy, helping to sort out

recommended instruments is the Extreme Adaptive

the structure and formation history of the Milky

Optics Coronagraph (ExAOC), which would probe

Way. The GSC recommended that the observatory

very close to bright stars in search of planets in

proceed with these three instrument concepts by

orbits similar to those in our own solar system.

funding conceptual design studies for HRNIRS and

The second instrument recommended by the GSC

ExAOC, and a feasibility study for WFMOS. In

December 2003 - Observatory issues several Announcements of Opportunity to the community for Aspen instruments and capabilities.

Multi-Object Spectrograph (WFMOS), which

May 2003 - Strategic Planning Group requests that the Observatory work closely with partner communities to develop Key Questions. Result: Several mini-workshops initiated throughout the partner countries that involve hundreds of astronomers. November 2002 - Strategic Planning Group established by Board.



Gemini Observatory

June2005 addition to these top-ranked instrument concepts,

allocate observing time, and distribute data in

the committee also recommended exploring the

support of large campaigns. The ideas described

feasibility of the Ground Layer Adaptive Optics

below have been presented to the Operations

(GLAO) system on Gemini, which would make it

Working Group and GSC. As of this writing, the

possible to study the first galaxies in the universe

partners are providing feedback on these ideas.

and the epoch of reionization. GLAO would make

In May 2005, the Board will be asked to approve

many observations more efficient by improving

a preliminary strategy for use with the almost-

delivered image quality much of the time.

complete near-infrared coronagraphic imager (NICI) based on the ideas presented here.

The Gemini Board expressed its support for the Aspen science goals and will fund construction

Campaign Science

of as many Aspen instruments as possible. The Board’s commitment on the part of the

The proposed Aspen instruments are much more

international Gemini partnership was expressed as

specialized than the previous generation of general-


purpose instrumentation. They are optimized to address particular science questions. The science

We see the Observatory establishing a leadership role in

questions demand both specialized equipment

a global effort to define, address, and solve compelling

and large allocations of time to conduct extremely

scientific questions. The answers to these questions will have

targeted surveys. The ExAOC planet search will

a fundamental impact on our view of the Universe and our

target as many young nearby stars as possible

place in it.

to build up a large enough sample of planets to understand how such worlds form. This survey

The Board re-affirms its endorsement of the scientific goals

will take approximately 200 nights over a period

of the Aspen program, as expressions of the aspirations of

of about three years. The HRNIRS planet search

the entire Gemini community. The Board renews its pledge

will take a similar number of nights, with repeat

to actively pursue efforts to find the resources to enable the

observations spread over several years. The

Gemini community to address these scientific issues in a

WFMOS instrument has two large observing

highly competitive environment.

programs: the dark energy survey which will take 200 nights to complete, and the galactic archaeology

Along with this statement of the Board’s resolve

survey of the Milky Way which could take

came a charge to the observatory, which is to;

more than 300 nights. These are truly staggering allocations of observing time, orders of magnitude

remain an agile, responsive, innovative organization

larger than the typical Gemini proposal of 10 to

maximizing its use of the strengths of the partnership;

12 hours. It is therefore crucial that the Gemini partnership plan carefully and early to ensure

initiate and continually strengthen partnerships with other


institutions in the pursuit of scientific aspirations we hold in common with the global community; and

Construction of the near-infrared coronagraphic imager (NICI) is nearly done, with scheduled

explore new modes of astronomical observation and to lead

delivery to Gemini South later this year. NICI fills

in the evolution of necessary cultural, managerial, and

an important gap in our planned planet imaging

institutional changes.

surveys between current Altair adaptive optics studies, which are limited by the time-varying

The last sentence is important. The Gemini

structure of the point-spread function (PSF),

Observatory has begun the process of defining

and ExAOC, which will open up a dark hole

what cultural, managerial and institutional changes

in the PSF that extends inward to within 0.1 to

will be necessary to meet the Aspen science goals.

0.2 arcseconds from the primary star. NICI fits between these two in both performance and in the

The rest of this article describes a first attempt

size of the survey needed. For NICI to succeed

to construct a new way to select science teams,

at discovering planets, it will have to observe a


June2005 The near-infrared coronagraphic imager (NICI) fully assembled.

Second, we must attract and encourage the creative scientists from our communities to participate in the campaigns. It is crucial that Gemini not put in place administrative roadblocks that would discourage or disenfranchise the creative members large enough

of the Gemini partnership who would otherwise

number of stars, and

take the initiative to achieve the Aspen goals.

it should do it before ExAOC (or a competing instrument at VLT) replaces it in three to five years.

Third, access to Gemini observing time and data

A NICI survey is likely to take about 50 nights a

must be awarded on a competitive basis. The

year for the next few years, since a sample of at

observing time must go to the best science, as

least 100 stars must be observed, some of them two

recognized by a group of objective peers.

or more times. NICI therefore offers the perfect test bed for a new Gemini strategy for planning and

Fourth, there must be balanced and inclusive

executing a well-defined scientific campaign.

partner participation. The Board has expressed a “super-national” perspective by endorsing the Aspen

The Top-Level Principles

science goals, and this attitude of collaboration must be preserved as the teams are formed and

Before we embark on the Board-mandated

the time allocated. No partner will be excluded

mission of creating the new procedures to manage

from participation, nor should any partner avoid

large Aspen campaigns, it is worth reflecting on

responsibility for providing a share of the resources

the strengths of the way things are done now.

needed to make the Aspen science goals possible.

The following principles were presented to the Operations Working Group in February of this

Fifth, the Aspen campaign teams must be open

year, where they were endorsed.

to all members of the Gemini community. Just as the responsibility for providing resources for

First, any new system of allocating observing time

the campaigns are shared among the partners, the

must make it possible to achieve the Aspen science

benefits must also be shared. Campaign science data

goals. If we fail in this aspect of the program, the

should be shared with astronomers from all of the

large amount of money and effort expended will


have been wasted. Finally, Gemini must maintain access to the Aspen


instruments for other science. We do not know

Gemini Observatory

June2005 what exciting new observations will be proposed

been funded to support large international teams,

in the coming years. We must maintain access to

additional support will have to be found in the

Gemini for those whose science is not directly

Gemini partner countries. Teams will have to find

related to the large campaigns. The observatory is

resources to support students and postdoctoral

a versatile resource that must continue to meet the

researchers, hold team meetings, build data

needs of the entire Gemini community.

processing pipelines, release data packages, and provide management oversight.

Allocating Time to Large Science Campaign One area of concern is how to include astronomers


who weren’t part of the selected team. Some way Setting aside the hundreds of nights necessary to

to add a limited number of additional people with

achieve the Aspen goals requires a new strategy.

useful expertise and resources is being explored. It

If we naively relied on the current time allocation

is important that teams be appropriately sized and

process for the campaigns, each team would be

managed to maximize their scientific productivity.

faced with the multiple jeopardy of nine national

The rules and conditions for adding people will be

time allocation committees (TACs) twice a year

stated in the Announcement of Opportunity, and

for several years. While this would preserve

may be different for the various campaigns.

competition and partner share balance, it would be unmanageable for several concurrent campaigns,

Data Distribution and Proprietary Time

and might well result in failure if the TAC allocates insufficient time for a particular survey. It would

The policies regarding data distribution should

also discourage potential principal investigators and

be defined to ensure that the Aspen science goals

suppress creativity.

can be met, and that the teams and communities receive the benefits of that success in return for

It is important that the Director allocate the

their hard work and financial support. In the case

campaign time before the regular International TAC

of the planet searches, which require follow-up

time to guarantee that the Aspen science goals are

or multi-epoch observations, this implies that the

met. The Director will decide how much time is

entire survey should be proprietary to the science

needed for a given campaign after consulting with

team until the planets have been identified and

the outside experts that serve on the different

confirmed. It is important that the team’s efforts

Science Working Groups and the GSC. The Board

not serve only to confirm someone else’s discovery.

will then approve the proposed allocation.

On the other hand, the WFMOS data sets could profitably be released during the survey so that

Getting the most out of the huge data sets that

other projects could go forward.

will inevitably result from hundreds of nights of observing will require a great deal of effort and

It is also important that the Aspen survey data

organization. Given the very targeted nature of the

sets be available to the rest of the community

Aspen science goals, specialized international teams

(and world) to allow other avenues of research.

must be formed. Following the principles above,

Such programs will increase the benefit of the

the Gemini Director will issue an Announcement

survey to the Gemini community. It is therefore

of Opportunity for a Board-approved campaign

important that the data be released in a way that

science program. Teams will organize freely and

will maximize its usefulness and impact. The data

submit proposals to Gemini. Proposals will identify

should be distributed using the Gemini Science

resources the team can provide (scientific expertise,

Archive, made available in discrete, well-calibrated

additional observing time or data sets, students and

and documented data sets. The data must be easy

postdoctoral researchers, software, facilities, etc.).

to use and understand. It is also expected that the

An independent committee of experts, perhaps

Aspen data sets will be consistent with and a part

incorporating international TAC members, will

of the Virtual Observatory.

assess the proposals. Given that Gemini has not


June2005 Maintaining Time for PI Science

announcement to the Gemini community by June. The team will be selected at the same

How do we make sure that time is made available

time as the regular 2006A proposals. During

for traditional principal investigator (PI)-led science

NICI commissioning and its science verification

when the Aspen campaigns are in progress? First,

observations, the science team will be involved

the Director will announce ahead of time the

so that observing techniques and data processing

number of nights allocated to a given program

software can be tested.

over a set number of years. The number of nights will be based on the recommendation of the GSC

We are proposing a very tight schedule for a start

and the Science Working Group for a particular

to the NICI campaign. It is important, however,

instrument, and will be approved by the Board.

to use as many of the ideas developed for the

The Director will balance the needs of the Aspen

Aspen program as possible so that problems can

campaigns with the need for PI science time.

be solved and details refined. With a successful

The Board will provide their input in balancing

NICI program under way, we will be prepared to

the competing needs of the campaigns and the

start organizing the Aspen science teams as soon as

communities’ access to observing time. Since

instrument construction begins.

the campaign size is defined before the original announcement, the approximate fraction of time

The Gemini community is embarking on a mission

available for PI science will be made public well in

of discovery to better understand the universe and

advance of the start of a survey.

the conditions that make life possible. Meeting the lofty aspirations of astronomers as expressed

Testing the Waters: The NICI Planet Search

in Aspen will require expensive and complex instruments. It will also require a new way of

The NICI planet survey will be used to develop

thinking about how we go about doing science

and test procedures to be used for the Aspen

at Gemini. We look forward to developing a

campaigns. NICI will be commissioned and

campaign strategy for NICI that will prepare the

available for science observations by semester

way for the exciting Aspen science campaigns to

2006A, and we need to start the campaign as


soon as possible to take advantage of the window of opportunity that will close with the arrival of ExAOC and other similar competing instruments three to five years later. To start a NICI survey in 2006A, we need to organize a science team in late 2005. We plan to make a proposal solicitation A companion object about 65 times the mass of Jupiter orbits the star 15 Sagittae in this artistʼs illustration. This brown dwarf is more massive than a planet and glows from its own internal heat source, but is too small to produce energy by nuclear fusion like a star.


Gemini artwork by Jon Lomberg

Gemini Observatory


by Charles Telesco

Exploring the Inner Disk of

β Pictoris with T^ReCS


he debris disk around the main sequence

As noted in other contexts though, pathological

star Beta Pictoris (β Pic) has remained

behavior can reveal the clearest, purest

near the center of our quest to understand

manifestation of basic phenomena, and in that

the early evolution of planetary systems since

sense β Pic is proving invaluable. In addition, its

its discovery by IRAS in 1984 and initial optical

disk appears nearly edge-on, which substantially

follow-up imaging from the ground shortly

increases the surface brightness. At nearly 63 light-

thereafter. Scattered-light or thermal-infrared (IR)

years (19.3 parsecs) it is realtively nearby, with

images of β Pic are shown at virtually every

the visible disk subtending nearly 155 arcseconds.

conference or workshop on this topic to illustrate

Thus, both scientifically and practically, β Pic is a

our modest but growing knowledge about these

compelling target. Like the Orion Nebula or the

fascinating objects. The irony of this attention

Galactic Center, β Pic is invariably one of the first

is that β Pic is actually an extreme example of a

objects to be observed when a new astronomical

debris disk, with a mid-to-far IR excess second

instrument is commissioned in the southern

only to HR 4796A (another star- and dust-disk

hemisphere. This was the case shortly after the

combination that lies about 230 light-years away

thermal infrared instrument T-ReCS became

in the constellation Libra) and a uniquely large

available on Gemini South in semester 2003B.

diameter of at least 3,000 astronomical units (AU).

Top: Image of β Pic disk at 18.3 microns (SW to right). Lower right: Detail of emission in SW wing at 18.3 microns. This tightly confined ʻexcessʼ emission is likely the result of a recent collision between two large planetesimals. Lower left: Color composite of disk using all images shown on page 13.


June2005 Extensive optical imaging of the scattered starlight

distance from the star. For some grain materials,

from β Pic has shown that the disk beyond about

the corresponding typical particle sizes would be

100 AU from the star appears asymmetric in a

0.1-0.2 micron in the southwest and 1 micron in

variety of ways. Most dramatically, the northeast

the northeast. Thus, it appears that the southwest

wing of the disk is unambiguously brighter, longer,

wing of the central disk is brighter than the

and thinner than the southwest wing, and there is

northeast because there is this “extra” population

an S-shaped warp of the entire disk. In addition,

of smaller dust particles.

Hubble Space Telescope images showed that there is a structural discontinuity near 100 AU, such

A big clue to the origin of those grains is the shape

that the disk within this boundary (the “central

of the infrared-emitting region in the southwest.

disk”) appears tilted by several degrees relative to

The emission is dominated by a bright clump

the larger-scale disk. Probing the emission from

centered at about 52 AU in projection from the

the dust particles in the central disk at thermal-

star. The clump appears to be spatially resolved

IR wavelengths within 10-20 AU of the star has

along both the major and minor axes of the disk,

shown that there is a central lower-density region

having an intrinsic scale size of about 15-20 AU.

(the central “hole”) within 50 AU of the star, and,

If our estimates are correct, the dust particles that

in marked contrast to the outer disk, the southwest

we see in this clump are small enough that stellar

is brighter and longer than the northeast wing in

radiation pressure should remove them from the

the 10- to 20-micron wavelength region. Several

system on an orbital time scale of several hundred

studies have shown that some of these properties

years. However, to appear so spatially confined

may result from perturbations associated with

they must be either continuously generated or

planets on inclined orbits.

produced in a single event within the last 50-100 years. Continuous generation of particles through

In December 2003 and January 2004, our team

the grinding collisions of planetesimals that are

used T-ReCS to image β Pic’s central disk in

resonantly trapped and confined by a planet is

five passbands spanning the 8- to 25-microns

one possible source of the clump, or we may be

region at close to the diffraction limit (5 AU at 10

witnessing the recent dramatic release of particle

microns) of the Gemini South 8-meter telescope.

fragments associated with a catastrophic collision

These results, presented in the January 13, 2005

of two planetesimals in the central disk.

issue of Nature, held several surprises. Our images confirmed the previously observed mid-

These scenarios and plausible alternatives must

IR brightness asymmetry, but they also showed

still be explored in detail for β Pic, but it is

clearly that the magnitude of the asymmetry

interesting to speculate a bit more before all the

depends strongly on wavelength. At 8.7 microns

hard facts are in. The collisional hypothesis has

the southwest wing is nearly twice as bright as

special appeal, because with an age of 10-20 million

the northwest wing, whereas at 25 microns the

years, β Pic is quite young, and we know that

asymmetry is negligible. This finding implies

planetesimal collisions were fundamental to the

that the infrared-emitting dust particles in the

planet-building process in our solar system during

southwest wing are different in some way from

its first few hundred million years. Indeed, it

those in the northeast.

is thought that the Moon formed in just such a collision. The dust particles that we actually detect

One simple interpretation of this finding is that,

in the bright clump in β Pic would constitute a

on average, the southwest wing particles are hotter

sphere about 100 kilometers in diameter. Larger

and therefore smaller than those in the northeast

fragments probably containing most of the mass of

wing. We estimate that the “characteristic” particle

the original planetesimal would also be produced,

temperature at a projected distance of 52 AU in the

but they would be cooler and undetectable in the

southwest wing is 190 K, compared to only 140 K

mid-IR, so 100 kilometers is a strong minimum

at 52 AU in the northeast, both of which are hotter

size. A collisional breakup that resulted in

than the blackbody temperature of 70 K at that

fragment velocities of a few kilometers per second

12 Gemini Observatory

June2005 could account for the minor-axis clump size, and

For more information, see our paper, “Mid-infrared

radiation pressure acting on the dust for 50 years

Images of β Pictoris and the Possible Role of

may have elongated the distribution along the

Planetesimal Collisions in the Central Disk,” Nature,

major axis, as observed. A host of implications

Vol. 433, 13 January 2005, pp. 133-136

present themselves for observational testing of this or any other explanation for the structure of β Pic’s central disk. Spectroscopy and imaging photometry of the disk will eventually reveal its secrets.

Gemini mid-infrared images of Beta Pictoris as obtained with T-ReCS on Gemini South. Differences in the shape and strength of dust emissions within the disk can be seen as the observed wavelength changes. Note: The clump where the suspected collision occurred is to the right of the central white core at a distance of about 52 AU.



by Bob Blum

Searching for Accretion Signatures in Massive Stars The NOAO Phoenix spectrometer on Gemini South (above)


ow do massive stars form? Simple

are providing capabilities which allow us to look

expectations based on the observed

much farther away for these rare massive stars as

masses and lifetimes of massive stars

they form and evolve in their birth environments.

dictate that the time scale for formation must be

Adaptive optics, mid-infrared imaging, and infrared

much shorter than for low-mass stars, and that

spectroscopy are all beginning to play a role in the

the average mass accretion rates must be high,

quest to understand how massive stars begin their

about 10-4 MSun per year. Detailed calculations

young lives and end up as optically identifiable O-

suggest the formation time scale for massive stars

and B-type stars through the process of shedding

is approximately 100,000 years. In contrast to their

the natal cocoon through strong stellar winds and

low-mass stellar siblings, which are more numerous

ultraviolet radiation.

and whose formation proceeds by distinct and


identifiable phases over longer timescales, high-

Presently unique among the 8- and 10-meter class

mass stars are rare. They form rapidly and are

telescopes, Gemini provides a high-resolution

obscured by overlying envelopes of gas and dust.

near-infrared spectroscopic capability through the

It is also unavoidable that during the formation of

NOAO Phoenix spectrometer (above). Phoenix

a high-mass star, a massive core will form first and

provides spectral resolutions of 50,000 to 75,000

begin evolving (converting hydrogen to helium)

through 0.34- to 0.17-arcsecond slits at wavelengths

while the larger protostar is still accreting material.

from 1 to 5 microns. My colleagues, Peter Conti

Despite the difficulty of discerning the massive

(University of Colorado), Augusto Damineli,

cores as they form, progress is being made on

Elysandra Figuerêdo, Cássio Barbosa (University of

both theoretical and observational fronts. New,

São Paulo), and I have been working on a program

large-aperture ground-based telescopes like Gemini

to study massive star birth in giant galactic HII

Gemini Observatory

June2005 (GHII) regions such as NGC 3576 in Carina. The

the two grating settings suggest the CO itself

combination of Gemini sensitivity and Phoenix

may be coming from a physically narrow region

spectral resolution has allowed us to observe a set

farther away from the central star and not from

of objects toward several GHII regions and search

a rotating disk. If true, this indicates that several

for kinematic clues to the circumstellar geometry

CO emission mechanisms may be at play in

of newly forming massive stars.

the circumstellar environments of MYSOs. The kinematics for this source suggest a model in

Clearly, investigations of the earliest phases of

agreement with the results of Gemini/OSCIR

massive star formation have centered around

imaging of the region done by Barbosa and

longer wavelengths that penetrate the large column

colleagues and published in the November 2003

depth of material surrounding massive protostars.

issue of The Astronomical Journal.

However, currently achievable angular resolutions and sensitivities restrict study to a relatively few

Our Gemini South/Phoenix observations clearly

nearby sources. The Atacama Large Millimeter

show strong evidence for an accretion process in

Array (ALMA) telescope under construction

massive stars. An independent investigation in 2004

in Northern Chile will rectify this situation,

by Bik and Thi comes to similar conclusions. All

complementing studies at shorter wavelengths

these objects are massive late O- or early B-type

made possible by telescopes such as Gemini. In

stars. They are found in clusters which harbor

the meantime, we can take advantage of the fact

more massive stars, but the more massive objects

that newly formed massive stars (or massive young

do not show accretion signatures. They may form

stellar objects (MYSOs)), while at the tail end of

differently, or be more efficient in dissipating their

the stellar birth process, can be used to search for

disks through the action of their strong ultraviolet

evidence of remnant accretion disks. Our ongoing

radiation and stellar winds.

survey of galactic giant HII regions in the near infrared (described in a paper published by Blum,

Our GHII survey is continuing. We recently

Damineli and Conti in 2001) combined with the

migrated our low-resolution spectroscopic

work of Hanson and Howarth (detailed in a 1997

observations to Gemini North, using NIRI and its

paper), has led to the discovery of a sample of

grism capability in order to search for accretion

MYSOs with potential disk signatures.

signatures in more distant star-forming clusters. For the first time, our team has identified a source

We observed each of the candidate MYSOs in ionized hydrogen wavelengths (Br-γ at 2.17

with a composite mid O-type (M~/>50 Msun) and CO emission spectrum. We are planning Phoenix/

microns) and in the carbon monoxide (CO) 2-0

Gemini South observations of this source to see

first overtone bandhead (2.3 microns). Both features

if it contains the tell-tale signature of rotational

were observed in each of the candidates at low

kinematics in its CO profile. Stay tuned!

spectral resolution. The CO bandhead has been particularly useful in studying the circumstellar geometry of lower mass YSOs since it is sensitive to the warm and dense gas lying close to the object (at small radii). Remarkably, the CO molecules survive even under the influence of the harsh radiation field provided by more massive central objects. In our initial report on four MYSOs, we successfully modeled the CO emission profile with Keplerian disks. An example of the CO emission and best fit rotating disk model is shown at right. Each of the objects was also observed in the 2.17 micron Br γ line. These lines are consistent with a disk or torus model, but in one case (source 48 in NGC 3576), the combined kinematics from

The CO 2-0 first overtone rotationalvibrational bandhead for source 268 in M17. The blueshifted shoulder of the profile and redshifted peak are characteristic of Keplerian disks. The smooth curve is a model for the emission profile arising from such a disk. The short vertical line marks the vacuum rest wavelength of the bandhead (2.2935 microns), and the inset shows the emission-line profile for a single line in the bandhead (v=2-0, J=51-50). The xaxis of the inset is in km/s and is one Phoenix grating setting in extent.



by Mike Barlow

Mid-infrared Dust Emission from Massive-Star Supernovae S

upernovae occupy a pivotal position in

The Survey for Evolution of Emission from Dust

astrophysics. This is due to their importance

in Supernovae (SEEDS) collaboration, composed

as standard candles and cosmological probes,

of astronomers from North America and Europe,

but also stems from the prime role they play in

obtained Spitzer Space Telescope time to search

determining the overall energetics, mass recycling

for mid-infrared dust emission from nearby

rate and heavy-element enrichment of galaxies. Astronomers have long suspected that supernovae may be important sources of recyclable materials in galaxies. In particular, massive-star supernovae may produce observable dust signatures over short time scales, and so are good targets for infrared imaging and spectroscopy. We used the Michelle infrared imager on Gemini North to focus on the environment around the Type II supernova SN 2002hh, located in the nearby face-on galaxy NGC 6946 (which lies about 20 million light-years away). Our observations were obtained as part of a dual-pronged approach to understanding the contributions of supernovae to


the interstellar environment.

supernovae hundreds of days after their outbursts, using the Infrared Array Camera (IRAC) and Multi-band Imaging Photometer for Spitzer (MIPS) instruments. In addition, we have accessed Spitzer Infrared Nearby Galaxy Survey (SINGS) Legacy data on a number of nearby galaxies that have been obtained with the same instruments, in order to search for mid-infrared emission from recent supernovae within these galaxies. One of the earliest SINGS observations revealed strong dust emission from the vicinity of the Type II supernova SN 2002hh. Figure 1 shows the IRAC 8-micron image of the region around SN 2002hh obtained on June 10, 2004. The region is complex

Gemini Observatory


Figure 1.

and, given the angular resolution of Spitzer at

A SINGS IRAC 8.0-micron image of a 30” x 29” region around SN 2002hh (pixel size: 1.1 arcseconds), obtained on June 10, 2004. Five sources were fitted, with the supernova marked as Star 1 and a bright foreground field star (2MASS 20344320 +6007234) marked as Star 2.

this wavelength (2.4 arcseconds), crowded-field deconvolution had to be used to fit the data. Five sources were identified, of which the brightest (Star 2) is a foreground 2MASS source (from the Two Micron All-Sky Survey) in our galaxy, and the next brightest (Star 1, with an 8-micron flux of 17 milliJanskys) is located at the position of SN 2002hh. Sources 4 and 5 were suspected to be an extended HII region located in a spiral arm of NGC 6946, with a projected separation of 3 arcseconds (90 parsecs) from SN 2002hh. It was clearly desirable to obtain mid-infrared observations at higher angular resolution in order

Figure 2.

to confirm the identity of Source 1 with the supernova. We received Director’s Discretionary Time to use the Michelle mid-infrared imager/ spectrometer on Gemini North to image the region around SN 2002hh at 11.2 microns and 18.5 microns. Figure 2 shows the region around SN 2002hh in the 11.2-micron Michelle image. With its ten times higher angular resolution, Michelle completely resolves SN 2002hh from its neighbors. Measurements relative to the 2MASS source (Star 2) confirm that strong mid-infrared emission originates from the position of the supernova. In addition, the Michelle data confirm that Star 4/5 is an extended source. The mid-infrared spectral energy distribution

Our radiative transfer modelling used a range of

(SED) of SN 2002hh matches a source that radiates

dust species and indicates that the observed mid-

like a 290 K blackbody with a radius of 0.1 light-

infrared SED can be fit by a dust shell of between

year or, alternatively, like a 225 K source with an

0.10 and 0.15 solar masses with a visual optical

inverse-wavelength emissivity and a radius of 0.5

depth of between 3 and 4. This corresponds to a

light-year. It would require 25 years for supernova

total gas+dust mass in the shell of at least ten solar

ejecta expanding at 6,000 kilometers per second

masses. This is similar to those estimated for the

to reach the latter radius, so we conclude that the

dust shells around the self-obscured M supergiant

dust must have originated from an earlier mass-loss

NML Cygni and the luminous blue variable AG

phase of the supernova progenitor.

Carinae, suggesting that the precursor of SN

The Gemini North Michelle 11.2 micron image of a 21.8” x 21.8” region centered on SN 2002hh (pixel size 0.099”), obtained on Sept. 26, 2004. Offsets in arcseconds from the position of the supernova are marked on the axes. A 3-pixel (0.3”) FWHM Gaussian filter was applied to the cleaned image. Both the supernova and Star 2 are easily detected in this image, while Star 4/5 is confirmed as an extended source.

2002hh may have been similar to one of these From early-phase near-infrared spectroscopy, Peter


Meikle and collaborators have estimated a visual extinction towards SN 2002hh of six magnitudes.

Our observations do not rule out the possibility

One magnitude of that extinction can be attributed

that significant quantities of new dust may have

to foreground extinction by our own galaxy.

formed in the ejecta of SN 2002hh itself, merely


June2005 that the observed infrared emission is dominated

supernovae as well as the contribution made by

by dust located much farther out than the

dusty outbursts to the overall dust enrichment

supernova ejecta. The relative numbers of self-

rate of galaxies. Gemini Michelle time has been

obscured supernovae are unknown. For many

allocated in Semesters 2005A and 2005B for further

galaxies supernovae with similar or even greater

mid-infrared follow-up studies of extragalactic

self-obscuration than the relatively nearby SN

supernovae for which Spitzer Space Telescope

2002hh could easily escape detection by optical

observing time has been allocated. The example of

supernova searches. This suggests that near-

SN 2002hh demonstrates the benefits of such a dual

infrared-based searches may be required in order


to determine the ratio of dusty to non-dusty Supernovae as Cosmic Dust Suppliers The violent explosions of massive stars as supernovae could well be major, or even dominant sources of dust particles in the universe. This idea is indirectly supported by the fact that many of the earliest-formed galaxies known are extremely dusty and luminous in the infrared. The evidence for this comes from the efficient detection of their redshifted dust emission at submillimeter wavelengths by the Submillimeter Common-User Bolometric Array (SCUBA) and other instruments. It seems that only massive stars can produce the observed dust over the short timescales implied, either during dusty mass-losing phases prior to the explosion, or else within the expanding and cooling ejecta from the supernova. Supporting evidence for dust formation by at least some supernovae comes from precise studies of the ratios of chemical isotopes that make up the grain inclusions found in meteorites. Many of these inclusions exhibit isotopic distributions that differ significantly from those found in the Sun and Earth and have consequently been labeled as “presolar” grains, or “stardust.” Some of these grains have isotopic distributions characteristic of the r-process that operates inside supernovae and is responsible for the synthesis of heavier elements during such an event. The first direct observational evidence for the formation of dust by supernovae came from studies of SN 1987A in the Large Magellanic Cloud. Mid-infrared emission from warm dust grains that had formed in the ejecta from this supernova was evident from about 500-600 days after the outburst. The same dust produced about 0.6 magnitudes of optical extinction at this time, as deduced from red-blue asymmetries that developed in the profiles of optical emission lines. The red-shifted emission from the far side of the ejecta suffered more extinction than the blue-shifted emission from the near side. Models confirm that newly formed dust in supernovae ejecta should become detectable at mid-infrared wavelengths within 1-2 years of outburst. They also prove that the mid-infrared is particularly suitable to trace the onset of dust formation in supernovae and to determine the amount of dust formed. Supernova 1987A was an unusually close one and mid-infrared studies of other extragalactic supernovae have not been possible until recently. With the advent of ground-based 8-meter telescopes equipped with sensitive mid-infrared instruments, such as Michelle and TReCS, and the launch of the Spitzer Space Telescope, sufficient sensitivity has become available to embark on searches for mid-infrared emission from supernovae in nearby galaxies.

The SEEDS collaboration’s work on SN 2002hh included astronomers from University College London; the Space Telescope Science Institute; the Gemini Observatory, Louisiana State University; the University of California at Berkeley; the University of Hertfordshire; the University of Arizona; the Kapteyn Astronomical Institute; and the University of Manchester.

18 Gemini Observatory


Digital Star Trails above Gemini South

by Peter Michaud

Gemini Legacy Imaging More Than Just Pretty Pictures


he Gemini telescopes are exquisite scientific

members of the public. Ultimately taxpayers fund

instruments capable of producing some

our research and sharing our “visions” is a critical

of the most detailed views of the cosmos

element of our work.

ever obtained from the surface of our planet. This capability is already providing extremely high-

To address this crucial aspect of our mission,

quality scientific data (spectra and images) that

Gemini has begun a campaign called Gemini

will be reflected in the legacy of scientific papers

Legacy Imaging. The primary driver for this

published over the next 20 years and beyond.

ongoing program is to produce unique and aesthetically pleasing images from the Gemini

As demonstrated by the Hubble Space Telescope

telescopes that will appeal to the public and

Heritage project, there is another legacy that is

illustrate our capabilities to a broad audience (that

possible for an astronomical observatory. It is one

also includes scientists).

that presents stunning images showing our universe as only a large telescope can do. This legacy is

The development of this project has been under

important because it conveys the scientific potential

way since the Gemini telescopes completed

and success of an observatory like Gemini to the

commissioning, but the formal adoption of the



This image of NGC 2467 is a recent addition to the Gemini Legacy image collection. It provides an incredibly detailed view of a star formation region that lies about 17,000 lightyears away in the constellation Puppis. This image was produced using the Gemini MultiObject Spectrograph at Gemini South on December 5th, 2004. Making observations to produce striking astronomical images like this one is a very different process than acquiring data for science. To combine the best interests of science and public outreach, a concerted effort is being made to acquire all public image release data with the same speciďŹ cations as any research data set. This will allow future scientiďŹ c use of imaging placed on the Gemini Science Archive soon after the release of a public relations image (usually within two months of data acquisition).

20 Gemini Observatory


Gemini Legacy Image / NGC 2467 / T. Rector, University of Alaska Anchorage


June2005 program as a part of the observatory’s operations

In all instances, data obtained for Gemini Legacy

is more recent. Philosophical commitment from

Imaging will be made public on the Gemini

the highest level of the observatory administration

Science Archive within two months of in-house

and oversight was necessary to make this program

data reduction. All data will be fully calibrated

work. Part of that commitment is in the form of

and whenever possible, photometric conditions

time to do the required observations. A nominal

will be used (although they are not a requirement

observing block of two hours per telescope per

for Legacy imaging). Already several data sets

semester has been committed as part of the

have been made available on the Gemini Science

Director’s Discretionary allocation. Additional

Archive and links have been added to the Gemini

time can be made available for special projects,

Image Gallery for downloading the data.

especially those considered targets of opportunity or those with additional scientific interest.

While aesthetic considerations are the primary drivers of this initiative, a process has also been

Within these constraints, an extensive database

developed to allow targets with scientific potential

of potential targets has been selected for imaging

to be proposed. This allows for consideration

(primarily with the GMOS instruments). Each

of targets that might otherwise be ignored,

one has been run through the Phase II observing

especially if there is a strong scientific basis for

process, evaluated for appropriateness, and then

obtaining imaging data and it is likely to result

grouped into a “primary” and “secondary” band

in a striking image or significant illustration of

for prioritization. Each semester a selection of

Gemini’s abilities. Examples include adaptive optics

primary-band targets will be activated over a range

imaging, Integral Field Unit data-cube animations,

of hour angles so that they will be available at

infrared imaging, and time-dependent targets of

each telescope throughout a given night. This will


allow flexible and timely insertion into observing windows for minimal impact on nightly schedules.

The Gemini Legacy Imaging project has already

We anticipate that this arrangement will allow

produced a number of stunning images that are

for extremely efficient use of observing time since

available on the Gemini Image Gallery. If you

these relatively short programs (most are less than

haven’t looked lately, take a look at the Image

60 minutes total) can be inserted into the queue

Gallery and download the new screensaver or

with minimal impact. Seeing requirements for all

posters to help share the Gemini legacy.

Legacy imaging has been set to ~0.5 arcseconds to assure optimal image quality and accurately

The Gemini Image Gallery

demonstrate Gemini’s capabilities. NEW – Make Your Own Gemini Screen Saver The latest version of the Gemini Virtual Tour (available mid2005) includes a interactive screensaver maker for MacIntosh ® or Windows ®, that allows you to select only the images you want and customize them for your display. This is only available as part of the Gemini Virtual Tour CDROM which can be requested at: geminivt@gemini. edu

In an effort to make our images easily accessible to anyone, the Gemini Image Gallery (www. has grown significantly over the past six months to include a multitude of new resources. These include: •

Full-resolution Gemini Legacy Images

Electronic resolution image for displays and digital wallpaper

Large print-ready PDF posters

Broadcast-quality video

Multiple screen resolution MacIntosh®/ Windows® Gemini Screen Savers

22 Gemini Observatory


by Jason Kalirai & Harvey Richer

Gemini Tackles Key Problems in Stellar Evolution CFH12K Image / NGC 2099 / Canada-France-Hawai‘i Telescope


ne of the key obstacles that astronomers

Recently, our team has begun to place strong

face in developing a complete

constraints on this fundamental relationship. By

understanding of stellar evolution stems

combining data from three telescopes on Mauna

from our poor knowledge of the initial-final mass

Kea in Hawai‘i—Gemini North, Keck, and the

relationship. This correlation links the mass of a set

Canada-France-Hawai‘i Telescope (CFHT)—we

of stellar remnants—the so-called white dwarfs—to

are now able to accurately determine how

the mass of their progenitors (at a point in time

much material intermediate-mass stars will

when they were burning hydrogen as main-

lose throughout their evolution. In a two-

sequence stars).

phase program, we first imaged a large sample

Figure 1.

This image of the rich open star cluster NGC 2099 (above) was obtained with the CFH12K mosaic camera on the Canada-FranceHawaii Telescope on Mauna Kea.


June2005 of open star clusters using wide-field imagers

Spectral features of 16 white dwarfs observed

on a 4-meter class telescope. In the second

using this approach in the rich open star cluster

phase, these observations were used to feed the

NGC 2099 are shown in Figure 2. Rather than

candidate objects to 8- and 10-meter telescopes for

showing the full spectrum, each panel presents

spectroscopic follow-up. The results of the current

the hydrogen Balmer absorption lines for each

study are described in our paper “The Initial-Final

white dwarf. Hβ is shown at the bottom of each

Mass Relationship: Spectroscopy of White Dwarfs

panel, with subsequent higher order lines on top.

In NGC 2099 (M37)” co-written with D. Reitzel, B.

In the case of hydrogen atmosphere white dwarfs,

Hansen, R.M. Rich, G.G. Fahlman, B. Gibson, and

these are the only features present in their spectra

T. von Hippel, and published in the January 10,

(a consequence of the large gravity in these stars

2005 issue of The Astrophysical Journal Letters.

which causes other elements to “sink”). By fitting the morphology of each of these absorption lines

Figure 2.

Spectral line profiles of the hydrogen-Balmer (Hβ) lines of 16 white dwarfs in the open star cluster NGC 2099 with model fits in red. Each panel shows the Balmer lines for one star, with Hβ at the bottom and higher order lines at the top. All spectra were acquired with Gemini/GMOS and Keck/LRIS. The models and spectral fitting routines were provided by Dr. P. Bergeron at the Université de Montréal.


The CFHT Open Star Cluster Survey, which

to theoretical model white dwarf spectra of known

was the basis for phase one of our project, is a

pressures, temperatures, and masses (shown in

deep, wide-field imaging survey of rich, open star

red on Figure 2) it is possible to derive these

clusters in the Milky Way galaxy. These clusters

parameters for each star. For this, we used the

typically contain several thousand stars and range

white dwarf models, and spectral fitting routines

in age from a few million up to several billion

devised by Dr. P. Bergeron at the Université de

years old. The survey data came entirely from


the 3.6-meter CFHT telescope, equipped with the CFH12K wide-field image mosaic camera. (Current observations are taken with the new MegaCam 340 megapixel wide-field camera.) Figure 1 presents an image of the rich star cluster NGC 2099 (M37) as taken with the CFH12K camera. These imaging data make it possible to determine key properties for each of the clusters, such as their ages and distances. At the same time, the depth of the survey allows us to probe to very faint stellar brightnesses and identify the now-dead remnants of once-massive hydrogen-burning stars in the cluster. These white dwarf stars represent the key ingredient in understanding the initial-final mass relationship. In the second phase of our study, we used both the 8-meter Gemini and 10-meter Keck telescopes

Our results indicate that the mean mass of the

to obtain spectroscopic observations of potential white dwarf stars identified in the CFHT survey.

white dwarfs in NGC 2099 is ~0.8 solar masses, making them about 0.2 solar masses more massive

New instruments on these telescopes, such as the

than typical white dwarfs in the disk of our galaxy.

Gemini Multi-object Spectrograph (GMOS) on

This result is related to the age of the cluster

Gemini North and the Low-Resolution Imaging

NGC 2099. Comparison of stellar evolutionary

Spectrograph (LRIS) on Keck allow us to obtain

models (from F. D’Antona and P. Ventura at

high quality spectra of dozens of white dwarfs in

the Osservatorio Astronomico di Roma) with

each cluster simultaneously. The synergy between

the CFHT photometry for brighter cluster stars

these instruments, their parent large-aperture

indicates that the age of NGC 2099 is 650 million

telescopes, and the CFHT imaging observations

years. At this age, only those hydrogen-burning

provides for an unprecedented study of the initial-

stars with masses more than 2.5 solar masses have

final mass relation.

had enough time to evolve into white dwarfs (since more massive stars evolve faster).

Gemini Observatory

June2005 To quantify the initial-final mass relationship for these stars, we first determine the white dwarf cooling age for each star (Tcooling). This age represents the time that each white dwarf has spent traversing from the tip of the asymptotic giant branch (AGB) down to its present white dwarf luminosity. This age is easily determined from the white dwarf cooling models once the mass of the white dwarf is known. Next, we calculate the lifetime of the progenitor main-sequence star that produced the white dwarf (TMS). This is simply the difference between the cluster age (Tcluster) and the white dwarf cooling age (Tcooling), which can be expressed as: TMS = Tcluster - Tcooling Using models of main-sequence stars, the mass of the white dwarf ’s progenitor star can be determined from the progenitor lifetime. We now have final masses for all white dwarfs, and initial masses for the stars that produced them. In Figure 3 (top) we present this calculation in the form of an initial-final mass relationship. First, this study of NGC 2099 white dwarfs (filled circles) has almost doubled the number of data points on the relationship from all previous studies combined (shown as triangles). We also find a very tight correspondence between main-sequence stars with an initial mass between 2.8 and 3.4 solar masses and resulting white dwarf masses between 0.7 to 0.9 solar masses (Figure 3, bottom). The results clearly indicate that intermediate-mass stars will lose 70 to 75% of their mass through stellar evolution.

Figure 3.

The initial-final mass relationship is shown for stars in this work (filled circles) and all previous constraints (triangles). Also shown are several semi-empirical and theoretical relations discussed in Kalirai et al. (2005). (Bottom) A closer look at the white dwarfs that form the tight sequence with initial masses between 2.8 and 3.4 solar masses (all of which have lost 70-75% of their mass through stellar evolution).

Figure 4.

The colormagnitude diagrams (CMDs) of six rich open star clusters in the CFHT Open Star Cluster Survey are shown. The ages of the clusters vary over a large range: 8 billion years (NGC 6791), 2.5 billion years (NGC 6819), 1.7 billion years (NGC 7789), 650 million years (NGC 2099), 180 million years (NGC 2168), and 130 million years (NGC 2323). White dwarf candidates in each cluster are shown as darker points in the faint-blue end of each CMD.

With the success of this pilot study, we are confident that these techniques can be extended to other open star clusters of different ages. In Figure 4, we present the color-magnitude diagrams (CMDs) of NGC 2099 and five other rich open star clusters (all observed with CFHT). Possible white dwarf candidates in these clusters are shown as darker points in the faint blue end of the CMD. Given the different ages of these clusters, future studies will constrain a different range of progenitor masses than the present work. Therefore a complete mapping of the initial- to final-mass relation will soon be possible. For this, we have already begun a Gemini South program to look at the young cluster, NGC 2323.


June2005 White Dwarfs: Nature’s Stellar Diamonds Observations of stellar evolution and the resulting theory about how stars are born, live, and die give strong evidence that all stars expel a large fraction of their mass throughout their lifetimes. The most intensive phase of this mass loss takes place late in a star’s lifetime. When it nears the end of its life, a star’s central region contains less fuel for hydrogen burning. This core region contracts until the temperature gets high enough to start the nuclear fusion of helium. The structure of the star changes dramatically as hydrogen in the shell surrounding the core also begins burning. Depending on the star’s initial mass on the main sequence (where a star spends most of its hydrogen-burning lifetime), it then enters an unstable period where variations in its temperature, radius and luminosity occur. This can result in more structural changes in the star as it gently blows off its outer layers. For stars more than about seven times the mass of the Sun, the ultimate effect of these instabilities is a spectacular supernova explosion. Stars less than seven solar masses enter a short phase (a few thousand years in duration) that includes successive episodes of mass loss. Eventually all that’s left is a stripped stellar core of carbon and oxygen mixed with a dense, compressed gas of electrons that determine the structure of the remnant. This end product of stellar evolution is called a white dwarf. The properties of these stellar corpses are fascinating because of the curious nature of the highly compressed, highly conductive electron gas. In astrophysical terms, they form what is called a “degenerate gas” that cannot be compressed any further. It actually behaves like a solid. This incompressible quality of white dwarfs has led some to call them “the largest diamonds in the universe.” The white dwarf phase of a star can last for billions of years, and during this time the object does not generate energy by thermonuclear reactions. The energy that radiates is sustained simply by cooling, just as a hot piece of iron metal emits radiation as it cools. White dwarfs have an average diameter of about 10,000 kilometers, roughly the size of the Earth. However, a white dwarf’s mass is about half that of the Sun, which makes its density about a million times that of most common solid elements found on Earth. Planetary Nebula BD+303639 / Gemini Observatory + University of Hawai‘i

26 Gemini Observatory


by Stephanie Juneau

Did The Most Massive Galaxies Form First? A

Gemini-based study has led to the

GDDS team aimed to discover galaxies with stellar

discovery of a population of red and

masses similar to that of the Milky Way, which lie

massive galaxies in the “redshift desert,”

up to 9.9 billion light-years away (corresponding to

a zone in the universe that lies between a redshift

a redshift of z = 1.8). This very challenging task is

of 1.3 and 2 and looks back at an era when the

described in a series of papers stemming from the

cosmos was between a quarter and half of its

GDDS survey and available on the GDDS website

present age. This finding stems from observations

(URL on page 29).

done as part of the Gemini Deep Deep Survey (GDDS), an ultra-deep redshift survey looking out to a limiting magnitude of 20.6 in the K band and to 24.5 in the I band, targeting galaxies in this relatively unexplored region of the universe. The region in question was designated the “redshift desert” since massive galaxies in that redshift range lack strong spectral features at visible wavelengths. As a result, it proved extremely difficult to obtain spectroscopic redshifts until very recently. The

The spectra for the study were obtained using the Gemini Multi-object Spectrograph (GMOS). The team implemented the Nod-and-Shuffle observing technique that allows excellent sky light subtraction (to 0.1% accuracy) and deep exposures (with an integration time of 30 hours). This is a key element in obtaining high-quality spectra of faint galaxies. The main result from the GDDS is the

survey seeks to construct the largest mass-limited

discovery of a very significant population of red

sample of galaxies in this zone. In particular, the

and massive galaxies at 1.3 < z < 2, whose integrated


June2005 light is dominated by evolved stars. Along with

commonly used indicator of star formation activity

complementary studies from independent teams

is the oxygen spectral line at wavelength 3727

(most notable the VLT K20 survey), this was a

Ångstroms. The energetic UV photons emitted

major contribution to establish the abundance of

by young OB stars ionize the oxygen present

massive galaxies at z > 1.

in the surrounding molecular clouds. Upon recombination, oxygen atoms emit photons at 3727 Ångstroms, which creates a strong emission feature in the spectrum. Some example spectra of galaxies with a range of star formation activity are shown in Figure 1. For the GDDS galaxies, both star formation rate tracers were combined. Due to the stochastic nature of star formation events, the rate in individual galaxies can vary drastically, from hundreds of solar masses per year for a galaxy hosting giant bursts to essentially zero for passively evolving stellar populations. However, general trends can be revealed by averaging subsamples of galaxies.

Figure 1.

Example spectra showing a range of star formation activity in the GDDS galaxies. The images were obtained with the Advanced Camera for Surveys on the Hubble Space Telescope and show the morphology. The color of the image frame is keyed to the spectral classification: early-type in red, late-type in blue and intermediatetype (i.e. with both an evolved stellar population and ongoing star formation) in green. The sample shown here illustrates the ”mixed bag“ of galaxies observed by the GDDS.

Massive galaxy abundance declines at z > 1 but the rate of decline is slower than the simple models

The star formation history is assessed by

predict. The population remains significant at

computing the density of star formation rates at

z = 2. One question that arises is whether star

different epochs over cosmic history. With its

formation is more efficient in high-mass galaxies.

approximate mass selection, the GDDS allowed

Recent findings demonstrate that the evolution of

us to do this as a function of stellar mass. Figure

the rate at which galaxies form their stars changes

2 shows the comparison of our values of star

according to their total stellar mass. This new piece

formation rate densities with a compilation of

of evidence was obtained by measuring the star

results from 33 studies carried out between 1996

formation rates for a sample of 207 galaxies from

and 2004 (summarized by Hopkins in 2004). The

the GDDS. The classical “Madau-Lilly” diagram

key result is the clear dependence of the star

quantifies the overall star-formation rate of the

formation history on the galactic stellar mass:

universe over time. Combining this information

high-mass galaxies go through the peak of their

with the total stellar mass of galaxies is a new and

star formation activity at higher redshift. In a

promising approach in understanding how galaxies

previous study by Heavens and colleagues, and

assemble the bulk of their stellar content.

based on nearby galaxies, it was suggested that such a dependency with galaxy present-day stellar

One technique successfully used by the team

mass exists. There, the star formation history

to infer the total mass of stars in galaxies is to

was reconstructed from the analysis of the fossil

measure the red light (K band), which traces

traces of accumulated stellar populations. Here the

approximate stellar populations. Two strategies

equivalent relationship between formation history

were combined to calculate the rate of star

and stellar mass is tested directly at the epoch of

formation. First, the amount of star formation was


estimated from the light emitted in the ultraviolet


(UV) region of the spectrum. In normal galaxies,

The ratio of the total mass locked in stars to the

the signal at these wavelengths is predominantly

star formation rate (SFR) gives a timescale. This

generated by the newborn stars. Another

characteristic growth timescale (TSFR = M* / SFR)

Gemini Observatory

June2005 formation clearly supports “downsizing” in the star-formation rate (a mechanism proposed by Cowie in 1997), where the most massive galaxies form first and galaxy formation proceeds from larger to smaller mass scales. These findings were rather surprising with respect to the early predictions of the hierarchical galaxy formation model. According to that model, massive galaxies form from an assembly of smaller units. The most massive objects therefore form last. However, the GDDS study of the most massive and quiescent galaxies back to an era only three corresponds to the time required for a galaxy to assemble its mass, assuming that it forms stars at a steady rate. A simple interpretation of the mode of star formation emerges from the comparison of TSFR with the Hubble time TH(z), the age of the universe at a given redshift z (or epoch). Figure 3 illustrates this concept. If the values are equal, a galaxy has enough time ahead to produce all of its stars. At a given redshift, TSFR > TH suggests that the galaxies are in a declining or quiescent star formation mode. Conversely, TSFR < TH indicates they are going through a burst phase. Figures 2 and 3 demonstrate that the most massive

billion years after the Big Bang point toward early formation for a significant fraction of present-day massive elliptical galaxies. These findings could be reconciled with more modern versions of the hierarchical model if one posits a faster early star formation rate in more massive halos. Debate has ensued in the theoretical community as to the mechanism involved. This is another example where pushing the limits of the observable tightens the constraints that the models more details, see the original paper at http://www. v619n2/18987/brief/18987.abstract.html.

billion years of cosmic history. Intermediate-mass

This figure shows redshift-evolution of the star formation rate density (SFRD). The gray points show the compilation by Hopkins (2004). The sample of GDDS galaxies is split according to their stellar mass: low- (109 – 1010.2 Ms), intermediate(1010.2 – 1010.8 Ms) and high- (1010.8 – 1011.5 Ms) masses. The two indicators of star formation rates used are the [OII] 3727 emissionline (circles) and the rest-frame UV continuum at 2000 Ångstroms (triangles). SFRD data from the SLOAN Digital Sky Survey as reported by Brinchman (modified with our dust correction model and initialmass function) are split into the same mass bins and shown as low-redshift counterparts (squares).

must satisfy and adds a piece to the puzzle. For

GDDS data and publications are available at

galaxies formed most of their stars in the first three

Figure 2.

Figure 3.

The sample of galaxies is split according to their stellar mass: low(109 – 1010.2 Ms), intermediate- (1010.2 – 1010.8 Ms) and high- (1010.8 – 1011.5 Ms) masses. For every subsample, the ratio of the total stellar mass density to the star formation rate density gives the characteristic growth timescale. This timescale allows us to understand the mode of star formation when it is compared to the Hubble time (gray line).

objects continued to form their dominant stellar mass for an additional two billion years, while the lowest-mass systems have been forming over the whole of cosmic time. This view of galaxy



by Jean-René Roy & Phil Puxley

Recent Science Highlights Ultraluminous X-ray Source in the Spiral

Ultra-massive Black Hole at the Center of

Galaxy M101

Centaurus A

Ultra-luminous X-ray sources (ULX) with

The Gemini Near-Infrared Spectrograph (GNIRS)

luminosities greater than 10 ergs/second are

has unlocked new possibilities for the study of

suspected to be associated with intermediate-mass

central black holes in dusty galaxies. Julia D. Silge

black holes of 100 solar masses or more. During

of the University of Texas and her collaborators

the period of July 5-11, 2004 the source ULX-1 in

have used GNIRS to derive the central stellar

the spiral galaxy M101 (which lies at a distance of

kinematics from the signatures found in CO

about 5 megaparsecs, or 16 million light-years) was

(carbon monoxide) bandheads at 2.3 microns in

observed by Chandra to be in a brief period of

the galaxy NGC 5128 (Centaurus A). Located at

“high state.”

a distance of about 3.4 megaparsecs (11 million


light-years), this galaxy is an important object for A team led by K. D. Kuntz of the University of

our understanding of central black holes, galaxy

Maryland used GMOS-North on Gemini to target

mergers, active galactic nuclei, and the relationships

a faint source seen in images taken by Hubble

among these components of galaxy evolution. NGC

Space Telescope’s Advanced Camera for Surveys.

5128 contains large amounts of dust which hamper

The team obtained a spectrum on July 22 of the

optical spectroscopy, especially in the central

Vmag ~23.7 faint optical source co-located with the X-ray source. The GMOS data show spatially

regions that are critical for accurately measuring black hole mass.

unresolved He II 4686 emission that is associated with very high excitation. The line is consistent

Exploring orbit-based models to match the GNIRS

with an expansion velocity of ~600 kilometers per second. The fact that it appears as a point source

spectra (Figure 1), the authors favor a solution with

indicates that this high excitation optical emission

of 240 million solar masses. This is five to ten

is not of nebular origin but is coming from a

times higher than that predicted by the correlation

compact object. The team hypothesizes that ULX-1

between black hole mass and velocity dispersion.

is a high-mass X-ray binary, where the stellar wind

The GNIRS result is important because it confirms

of a B supergiant star accretes into a black hole

the earlier derivation by Marconi and colleagues in


2001 of a very high-mass black hole based on the

an edge-on model, and derive a black hole mass

kinematics of the gas. The motion of the gas could


be due to other mechanisms, like ejection, instead

Gemini Observatory

June2005 core. The authors propose that NGC 770 accreted a small gas-rich dwarf galaxy during a very minor merging event. If this scenario is correct, it represents one of the few known examples of merging between two dwarf-sized galaxies. The Presence of Steam Betrays a Low-mass Brown Dwarf Kevin Luhman (Harvard-Smithsonian Center for Astrophysics (CfA)), Dawn Peterson (University of

Figure 1.

Rest-frame spectra of seven example spatial bins (jagged black line) and for the template model stellar spectrum (red line) at different distances from the center of Centaurus A. Three CO bandheads are clearly seen as deep absorption features around 2.3 to 2.35 microns.

Rochester) and S. T. Megeath (also from CfA) used the newly commissioned GNIRS on Gemini South to determine the nature of the low-mass object OTS 44. The observations of the faint infrared object confirmed that it is a very low mass, of being orbital. Stars are much more robust tracers

newborn brown dwarf, and the lowest-mass one

of orbital movement. The deviation of Centaurus

found to date in the Chamaeleon I cloud complex.

A from the correlation suggests that its black

This is a nearby star-forming region located about

hole assembled first, before the main host galaxy

170 parsecs (about 550 light-years) away from us

component formed.

toward the southern hemisphere constellation Chamaeleon.

Counter-rotating Structures in Galaxies Astronomers have found many types of objects in Several Gemini observing programs are aimed

orbit around stars or as free-floating objects. These

at understanding the structure and kinematics

range from other full-sized stars like the Sun (some

of nearby galaxies, and establishing the role and

in binary star systems) to Jupiter-sized planets

effects of past mergers on galaxy morphology. One

(never directly imaged but inferred from radial-

GMOS-North study using long-slit spectroscopy

velocity spectroscopy).

by Peter Yoachim and Julianne J. Dalcanton of the University of Washington explored the rotation

From the strength of the water vapor (steam)

curves at and above the midplanes of the edge-

absorption band found in the spectra at around 1.5

on galaxies FGC 227 and FGC 1415. The team

and 2 microns, and in comparison to other low-

used the calcium triplet spectral lines around

Figure 2.

is the above-the-plane motion less than 25% of the

R-band image of the flat galaxy FGC 227. Solid lines represent the location of the slit for the mid and offplane GMOS longslit observations.

mid-plane velocity, but it is also counter-rotating

Figure 3.

850-860 nanometers to separate thin and thick disk stellar kinematics. In FGC 1415, the abovethe-plane rotational speed of the stars is 30 to 40% less than at the mid-plane. In FGC 227, not only

(Figures 2 & 3). The presence of a counter-rotating thick disk in FGC 227 rules out models of thick disks forming from monolithic collapse, or by heating of thin disks, and suggests direct accretion from infalling satellites. In a parallel study of the low luminosity elliptical

The rotation curve of the mid-plane (light shade and larger amplitude) and off-plane (dark shade and smaller amplitude) rotation curve; shaded areas show the full range of the fits for FGC 227.

galaxy NGC 770 using the integral field unit on GMOS-North, M. Geha and collaborators used internal stellar kinematics to find a counter-rotating



Figure 4.

Spectra of OTS 7 and OTS 44 are shown with CHSM 17173 for comparison. Regions of spectral water lines are indicated at bottom.

mass stellar objects, the team determined OTS 44 to be of very late spectral type M9.5 (Figure 4). They estimate its effective temperature at 2300 K. Using the flux in the H band and the object’s distance, Luhman and his team derived a bolometric luminosity of almost 800 times less than our Sun. Models of low-mass stars were used to infer the mass of OT 44 at about 1.5% the mass of the Sun and about 15 times the mass of Jupiter. Unlike older brown dwarfs (which have smaller sizes), this young object is relatively large with a radius of about one-quarter that of the Sun or 2.5 times that of Jupiter. First Measurement of Hyperfine Splitting Constant in an Astrophysical Object Of astrophysical interest is the isotopic ratio of A joint Chile-United Kingdom team led by Simon

aluminum (Al-26/Al-27). Al-27 is the stable isotope,

Casassus has used the NOAO-built Phoenix near-

while Al-26 is radioactive with a half-life of 720,000

infrared spectrometer at Gemini South to obtain

years, and is a signpost of recent nucleosynthesis.

high-resolution (R~75,000) spectroscopy of the [Al VI] 3.66-micron line region in the planetary

Because the ratio is poorly established, the

nebula NGC 6302 (the Bug Nebula). By modeling

of astrophysical processes ranging from nova

the multi-component line structure, the authors

detonations to cosmic-ray collisions in molecular

have been able to derive values for the electric


origin of Al-26 is currently assigned to a range

quadrupole constant in the [Al VI] transition and measure a reliable isotopic ratio for aluminum. This

NGC 6302 is the highest excitation planetary

is the first measurement of such a constant in an

nebula known. Its spectrum can be reproduced by

atomic transition in any astrophysical object (figures

ionization-bounded photoionization models with

on opp0site page).

a 250,000 K central star. The photoionized coronal lines in NGC 6302 are astonishingly narrow,

Most observed spectral lines arise from electronic

explained by a small expansion velocity. This and

transitions between well-separated energy levels.

its rich spectrum make NGC 6302 an ideal object

However some transitions take place between very

for the study and use of hyperfine structure as a

finely separated energy levels. A good example

diagnostic tool.

would be the hyperfine structure (HFS) lines due


to interaction between the internal magnetic field

The derived isotopic ratio of Al-26/Al-27 is less

produced by the motion of the electrons of an

than 1/33 in NGC 6302. This is the most stringent

atom and the spin magnetic moment of its nucleus.

upper limit on the relative Al-26 abundance in

This effect is well understood in terms of quantum

any astrophysical object to date. Although the

mechanics. While hyperfine transitions are common

measurement is not constraining enough to

in the radio range (the best known is the 21-

quantify the Al-26 production in Asymptotic

centimeter line of neutral hydrogen), at shorter

Giant Branch stars, that are the progenitors of

wavelengths atomic HFS has seldom been resolved

planetary nebulae like NGC 6302, the technique

in emission lines because the velocity separation

has been established. Tighter constraints require

between hyperfine components is proportional to

deeper spectroscopy of this planetary nebula, and

wavelength. This tends to broaden the lines and

extension of the analysis to other targets of low or

ignoring the effect can result in errors of ~50% in the inferred photosphere elemental abundances.

moderate expansion velocity.

Gemini Observatory


Figure 5.

A Hubble Space Telescope image of NGC 6302 with an inset showing the Gemini South acquisition R-band image (upper right) and the approximate location of the Phoenix spectrograph slit (small blue bar in inset).

Figure 6.

(a) Resulting observed Phoenix spectrum (red points) of July 30, 2003, and model line profile of [Al VI]: in grey, model profile of Al-26 had it been present at an isotope ratio of 1. In (b), coadded spectrum (red points). Light dotted-line does not include electric-quadrupole hyperfine splitting. The inclusion of the electric quadrupole hyperfine terms improves the fit (solid grey line).


Gemini North / Laser Assembly on Mauna Kea


Laser Guide Star Update T

his year represents a milestone for Gemini

for its ease of service. “Even during installation, we

Observatory’s adaptive optics (AO)

could replace modules very easily,” he said. “This

program as the laser guide star (LGS)

is cutting-edge technology.” The LGS system also

system comes online at Gemini North. The laser’s

includes a sophisticated system of beam transfer

“first light” propagation occurred on May 1-2, 2005.

optics, a laser launch telescope and a control

This is a big step toward the complete integration

system. On the first light test, the laser was

of the LGS system with the telescope, paving the

operated at about 7 watts, significantly less than its

way to LGS AO and multi-conjugate adaptive

maximum power. The delivered brightness of the

optics (MCAO) science.

laser guide star was measured to be of magnitude

The laser was built by Coherent Technologies,


9.3 and its size ~1.5 arcseconds.

Incorporated, of Louisville, Colorado. The 13-watt

First light proved the systems to be sound. More

solid-state diode laser fits into a relatively compact

commissioning work is planned in the coming

space on the side of the telescope. The system

months, with science commissioning with Altair/

has an extremely modular design, which Gemini

NIRI imaging and spectroscopy and Altair/NIFS

Senior Electronics Technician Kenny Grace praises

spectroscopy undertaken in the last quarter of 2005.

Gemini Observatory


The new Gemini solid-state laser resides inside a

Also shown above is a long duration (one-minute)

2.4- by 1.8-meter (8- by 6-foot) class 10,000 clean

image of the laser propagating on the sky during its

room on the side of the telescope (large blue box

second night of on-sky operations on May 2-3, 2005.

with white door in panoramic image above). The

Below this image, laser technicians and Gemini staff

beam travels through transfer optics to a “launch

don safety glasses to test the laser in the lab at the

telescope” mounted behind the secondary mirror

Hilo Base Facility. – Jean-René Roy

assembly. The operation of the laser is illustrated in a new animation available on the Gemini Image Gallery (a frame from the animation is reproduced in the image above on this page).



Gemini Scientist Receives Prestigious Young Researcher Award G

emini Observatory is proud to announce

first time in history, they successfully imaged a

that Inseok Song, assistant astronomer at

bound extra solar planet orbiting a celestial object

Gemini North has won the prestigious

other than our Sun (known as planet 2M1207b).

2005 Outstanding Young Researcher Award

Major mass media, including BBC and CNN,

(OYRA) given by the Association of Korean

reported this discovery and in-depth description

Physicists of America. The honor, bestowed

of this system has been published as an article in

during the American Physical Society’s spring

Astronomy and Astrophysics (2004, volume 425, page

2005 meeting in Los Angeles, includes a $1,000


cash award. It recognizes and promotes excellence in research by outstanding young ethnic Korean

Dr. Song is also studying sites of ongoing planet

physicists in North America who are working at

formation using both Gemini and Keck telescopes.

research-doctorate institutions, and industrial and

He recently found a young solar system analog

government laboratories.

believed to harbor a planet at 1 AU (the distance between the Sun and Earth). The system has

Dr. Song was nominated particularly for his

about 10,000 times more asteroidal material than

work in the search for Jupiter-like planets and

our own system currently has, and it seems to be

the formation and evolution of planetary systems.

undergoing a period of very frequent collisions

In making the award, the OYRA committee also

among planetesimals.

cited his contributions to the systematic search for extrasolar planets using the Hubble Space

Frequent collisions in this star system are

Telescope. He is a Principal Investigator on a large

remarkably similar to the lunar creation event in

HST program titled “Coronagraphic Imaging

our own Earth-Moon history. This discovery will

Search for Giant Gas Planets around Young Nearby

be published soon in a major refereed journal.

Stars.” Currently, eight Ph.Ds from three different countries are involved in this program. From

Dr. Song currently manages four research grants

observed data, Dr. Song and his team have found a

totalling a half-million dollars and has brought two

handful of very promising candidate planets.

post-doctoral researchers on board at Gemini. The Observatory is very proud of his achievements,

In addition to his HST planet search program,

and looks forward to many more exciting science

Dr. Song and his team are working with French

results from Dr. Song’s research group.

astronomers using the 8-meter telescope at the


European Southern Observatory in Chile. For the

Gemini Observatory


by Carolyn Collins Petersen

Tom Geballe: A Passion for Stars, Hoops, & Music P

ursuing the story of the universe requires

obtaining spectra and photometry of brown dwarfs

individuals with an uncommon passion

in order to understand their properties, including

for solving the mysteries of the stars and

the distribution of dust in their atmospheres

understanding the instruments used to observe

and the effect of convection on their chemical

them. Gemini North Senior Astronomer Tom

abundances,” he said. “There are about as many

Geballe is one of those people who finds

brown dwarfs as there are stars, but we

maximum enjoyment in whatever he

can only do spectroscopic analysis on

does. His commitment and enthusiasm

a handful of them—the brightest and

extend to everything in life, whether it’s

the closest. Gemini and its instruments

performing on flute as an accomplished

give us a chance to greatly expand our

musician, playing a fast-paced game

understanding of them.”

of basketball, or coaxing the best performance out of a telescope to

At the other end of stellar existence,

produce stunning spectra from a distant celestial

Tom is also fascinated by certain geriatric stars like

object. “There is much beauty in astronomy,” he

Sakurai’s Object, U Equulei, HD 137613, and V838

said. “However, I think the most sublime beauty is

Monocerotis. “They have unique infrared spectra

in the spectra of planets, comets, stars, interstellar

and apparently are behaving badly in their old

clouds, and galaxies. Spectroscopy reveals the most

age,” he said. “These objects are a lot of fun to

important secrets of the universe.”

observe because they are so strange.”

Tom’s research interests include brown dwarfs,

Between studying brown dwarfs and searching

those elusive low-mass objects in the stellar cellar

out misbehaving old stars, Tom peers deep into

that bridge the gap between Jupiter-like objects

interstellar space to learn more about the clouds

and the coolest stars. I am a member of a team

of gas and dust that exist there. “I and my


June2005 colleagues are still trying to find the reason for the

years, dabbles on guitar, and has performed with

remarkably high abundance in diffuse clouds of the

ensembles in Hawai’i, Seattle and Europe. He has

molecular ion H3+, which plays a fundamental role

served for 15 years as president of the Hawai’i

in the chemistry of interstellar clouds. “ he said.

Concert Society, an all-volunteer group that

“The answer to this puzzle is sure to teach us

presents music and dance events by professional

something important about the physical processes

artists. “That musicians as famous as the Tokyo

occurring in these clouds.”

String Quartet and the Los Angeles Guitar Quartet, and dance groups as well known as Pilobolus

Tom is a long-time resident of Hilo with more

Dance Theater and Smuin Ballet want to come to

than a quarter-century of experience observing on

Hilo and will make themselves affordable to us is a

Mauna Kea. He worked at the United Kingdom

wonderful gift for our community,” he said.

Introducing 4012 Geballe One of the rarer astronomical honors anyone can receive is to have an asteroid named after them. When asteroid 4012 was named for astronomer-musician Tom Geballe, he joined an even more select cross-section of asteroid honorees who happen to be musicians and composers. That group includes such luminaries as Antonio Vivaldi, the Beatles, Enya, Frank Zappa, Jan Sibelius, and another well-known astronomer-musican, William Herschel. What does 4012 Geballe look like? There are few images of it, but from light curve measurements and other observations, 4012 Geballe is now known to be an 11.2-kilometer (7-mile) diameter piece of rock orbiting at about 2.2 A.U. from the Sun. It’s a member of the Flora family of asteroids, whose orbits are influenced by Jupiter’s gravitational tugs. 4012 Geballe (also known as 1978 VK) has a V-band magnitude of 13.4, and was first spotted in 1934. Infrared Telescope (UKIRT) for 17 years, and spent

In 2004 Tom Geballe was honored by having

eight of those years as its director before coming

asteroid 4012 named after him. The IAU citation

to Gemini Observatory in 1998. This experience

recognized his outstanding expertise in infrared

gives him unique insight into the mountain, its

spectroscopy, stated:

difficulties, and what it can deliver under good conditions. According to Gemini North Deputy

T. R. Geballe is responsible for pioneering the first infrared

Director Jean-René Roy, Tom is the world’s

spectral classification system for substellar objects. Discoveries

foremost expert on thermal infrared spectroscopy

include H3+ in the giant planets and interstellar medium,

and a valuable asset to Gemini North. “Tom’s

non-LTE emission from Saturn VI (Titan), and countless

knowledge is hugely broad and deep,” said Roy.

IR observations of circumstellar and interstellar molecular

“Show him an IR spectrum, he will likely tell


you straight away what type of object it is, what problem there may be and how to solve it.”

It’s a fitting salute to an astronomer who has written or co-authored nearly 300 science papers,

Away from work, Tom’s lifelong love of basketball

and is described by friends and co-workers as

takes him to the nearest county gym every Friday

enthusiastic about his work and his art, a man

night, and he shares his passion for music with

who gets more energy from juggling multiple

wife Carole, who like Tom, is a flutist. They

projects, and thoroughly enjoys everything he does.

have two grown children, Anneke and Matthew,

One look at his record of accomplishments makes

both doing their graduate work on the mainland

it clear that the honor is well-deserved.

U.S. Tom comes from a large family of musical


and scientific performers. His father, the late

“Tom is one of the most stunning examples of

Ronald Geballe, was a professor of physics at the

a scientist combining commitment, passion and

University of Washington and played piano, his

easy-going manners,” said Jean-René Roy about his

mother Marjorie used to be a singer, and his uncle,

accomplishments. “He is always there to help. He

Theodore Geballe, is an emeritus professor of

is cool, patient, and joyful. It is just wonderful to

physics at Stanford. Tom studied flute for ten

have him on a team and to work with him.”

Gemini Observatory


by Carolyn Collins Petersen

Helena Vincke: A Can-Do Person at Gemini South T

he busy and complex environment of a

from Belgium. “He speaks perfect Spanish,” she

modern-day observatory requires staffers


to be flexible and clever, not just up on

the mountain, but also in administration at the

After migrating from Belgium in 1999, Helena

base facility. Gemini South Administrative Assistant

joined Gemini Observatory during construction of

Helena Vincke is one of those people that everyone

the southern telescope facility. She served as the

calls on with a challenging problem to

site manager’s administrative assistant,

solve. This friendly, engaging Belgian

and immediately fell in love with the

transplant greets co-workers with a

work. “It was fun because I had to walk

big smile and her “I can” attitude is

around with two radios and a hard hat

one of the first things her colleagues

and there were about 50 men working

mention. “Helena will never tell you

and I was the only woman,” she said.

that your problem is not related to her

“At first the men who interviewed me

department,” said Antonieta Garcia,

wondered if I would be able to handle it.

Gemini South’s Public Information and Outreach

But because I’m 1.76 meters (5 feet 9 inches) tall, I

Assistant. “Even if it’s not part of her job, she’ll find

think I made quite an impression.”

a friendly way to help solve your issue.” After two years of traveling up and down the A self-described “computer freak,” Helena taught

mountain, Helena joined the staff at Gemini’s

herself how to set up web pages, design PDF

La Serena base offices. “I was surprised at the

forms, and under the tutelage of Peter McEvoy,

workload here in La Serena,” she said. “With the

learned how to use Excel and other programs. “Her

move to the new building, having to coordinate the

contribution at Gemini South is just fantastic,” said

logistics, my purchasing experience came in handy.

McEvoy. “She knows how things work and she has

I think I was a good asset for our group knowing

a sincere dedication to helping people out. There’s

more about how things were organized on the

rarely a hiccup in the services she organizes.”


“My hobbies used to be reading, swimming,

Helena’s common sense, and what friend and

cooking, and aerobics. Then they were replaced

Gemini colleague Marie-Claire Hainaut calls her

by a lot of work and computer and administrative

“grounded nature,” are a big part of Helena’s

courses,” she said. Helena devotes much of her

valuable contribution to Gemini South’s team-based

spare time to her 9-year-old son Nando.

atmosphere. “No matter how big the obstacles, she always finds the strength to look ahead and make

Helena speaks English, Flemish, French, German,

the most out of every opportunity,” said Marie-

and Spanish fluently. She is especially proud of her

Claire. “She is a truly honest and sincere person, a

son’s growing language abilities since they emigrated

great coworker and a wonderful friend!”



by Frederic Chaffee, Matt Mountain & Hiroshi Karoji Subaru Keck


Mauna Kea observatories by Richard Wainscoat / IfA

New Capabilities, Cooperation atop Mauna Kea F

our of the world’s nine largest operational

UK, Canada, Australia, Brazil, Argentina and Chile

telescopes—the two Keck 10-meter

—who make up the Gemini partnership.

telescopes, and the Gemini 8.1-meter and

Subaru 8.2-meter telescopes—reside atop Mauna

In addition, all three provide observing time to

Kea, making the entire astronomy complex there

astronomers from the University of Hawai‘i.

the most powerful in the world. We have recently taken steps to assure that this entire “system” of

With such a huge constituency of worldwide

premier telescopes is used to maximum advantage

astronomers to serve, the demand for time on these

by the world’s best astronomers.

large Mauna Kea telescopes exceeds the supply by a factor of from three to five. Many requests by

Often viewed by non-astronomers as locked in

the world’s best astronomers for observing time at

fierce competition with each other, the reality of

Keck, Gemini and Subaru must be turned down

inter-observatory relationships is much different.

because time is simply not available.

Each serves a different community of astronomers. Keck observers come largely from the institutions

All of the major Mauna Kea telescopes have a suite

who fund the observatory—the University of

of basic instruments, which record the light from

California, Caltech and NASA.

celestial sources for analysis by astronomers. The majority of instruments at any modern observatory


Subaru serves the astronomers of Japan and

are of two basic types—imagers and spectrographs.

Gemini those of the seven countries—the U.S., the

Imagers take “pictures” of the universe—the

Gemini Observatory

June2005 magnificent pictures from the Hubble telescope are

world leader in discovering planets around other

familiar to people all over the world; even more

stars—in the last eight years some 70 new Jupiter-

spectacular images are beginning to be produced

sized planets have been discovered by Keck’s

by the Mauna Kea “giants.”

“planet-hunter” team.

Spectrographs spread out the light from celestial

However, planets smaller than Jupiter have so

sources into their component colors—providing

far eluded detection by HIRES. This newly-

the equivalent of a “fingerprint” or the “DNA

inaugurated capability should allow Neptune-sized

signature” of the celestial objects. It is through the

planets—one-third the mass of Jupiter–around other

study of spectra that astronomers learn most about

stars to be detected for the first time.

the nature of amazing constituents—planets, stars, galaxies, black holes, pulsars, gamma-ray bursters,

Recognizing that their respective observatories have

and the like–of the Universe.

developed unique capabilities that both of their user communities could exploit to produce new

In their determination to provide the best possible

and exciting discoveries, the Directors of Keck

facilities for the world’s astronomers Keck, Gemini

and Gemini recently reached a formal agreement

and Subaru have collaborated informally for many

whereby Gemini astronomers will be given access

years. Their Directors meet regularly to discuss

to HIRES on Keck and Keck astronomers to

common problems and concerns. Technical groups

Michelle on Gemini. This is the first formal time-

at each observatory meet regularly to exchange

exchange agreement between two Mauna Kea

ideas and share technical insights. Experts from


each observatory regularly serve on review and advisory committees for the other two. Equipment

In parallel development, Subaru and Keck recently

is often shared among observatories. We live in

arranged to exchange time when it was realized

and observe the same universe.

that a program already assigned time at Keck could better be done using Subaru’s “Suprime-

Recent advances in astronomical instrumentation

Cam,” and that Keck has a unique and powerful

have spawned an even closer relationship. In

spectrograph—DEIMOS—for which no equivalent

recent months, Gemini has begun commissioning

is available to Subaru astronomers. Thus a deal

a new instrument called Michelle—an acronym

was struck whereby Keck scientists will carry out

meaning “mid-infrared echelle” spectrograph. It

their research with Suprime-Cam at Subaru.

offers a unique capability available at none of the other large Mauna Kea telescopes. For example,

These first steps toward time exchange among

astronomers can explore the dust in proto-

Mauna Kea’s large observatories promise a new

planetary disks to find evidence of hidden planets

era of even closer collaboration, an era when we

or probe the dust produced by the very first stars

jointly seek to maximize the creative potential

in distant galaxies.

of all Mauna Kea astronomers, telescopes and instruments to produce evermore exciting

Keck itself has recently commissioned a unique

discoveries about our rich and infinitely varied

new capability for one of its “workhorse”


instruments—the High Resolution Echelle Spectrograph (HIRES). The new capability, many

Drs. Frederic H. Chafee, Matt Mountain and Hiroshi

years in development, will greatly increase HIRES’

Karoji are the Directors of the Keck, Gemini and Subaru

sensitivity to light of all colors, but especially to

Observatories, respectively.

ultraviolet light to which it has hitherto been all but “blind.” In the past, HIRES has been the

Reprinted from West Hawaii Today, September 29, 2004



by Larry Ramsey

Gemini’s Directorship Transition W

ith the recent announcement that

Director Search Committee, which is now underway.

Gemini Director Matt Mountain

This search committee will both accept and solicit

will likely fill the position of

applications and will develop a report that should

Director at the Space Telescope Science Institute

include a ranked list of director candidates. After

(pending NASA approval), a process has been

receiving the search committee report, the AOC-

initiated to assure the seamless transition of the

G will forward their recommendation through

Gemini directorship. As chair of the AURA

the AURA president to the AURA board. The

Oversight Committee for Gemini (AOC-G), I

final step in the selection process will be a

will be managing the search process for the new

recommendation to the Gemini Board for approval

director. I anticipate a smooth transition given the

of a new director during the latter half of 2005.

excellent state of the observatory and the strong management team that has been established under

It is the objective of this plan to minimize the

Matt’s leadership. Jean-René Roy will be the

impact on ongoing operations and keep Gemini

Acting Director of the Gemini Observatory during

at the forefront of optical/infrared astronomical

the transition period.

research. On behalf of the AOC-G, I would like to thank Matt and his team for all they have done

A search plan has been developed and approved

to ease this transition and make the next phase at

by the AOC-G. It includes the formation of a

Gemini a success.

42 Gemini Observatory


Recent Gemini Outreach Activities AstroDay 2005 in Hilo / by Gary Fujihara / IfA-Hilo

StarTeacher, Tom Chun conducting physics lab via video conferencing from Chile.

Journey Through the Universe 2005 / NASA scientist talks to Waiakea High School physics class

Students from Kamehameha Schools participate in Gemini video-conference physics lab


Gemini Observatory Northern Operations Center 670 N. A‘ohoku Place Hilo, Hawai‘i 96720 USA Phone: (808) 974-2500 Fax: (808) 935-9235 Gemini Observatory Southern Operations Center c/o AURA, Casilla 603 La Serina, Chile Phone: 011-5651-205-600 Fax: 011-5651-205-650

Gemini Partner Agencies:

United States

United Kingdom






Gemini Observatory is an international partnership managed by the Association of Universities for Research in Astronomy under a cooperative agreement with the National Science Foundation.

Gemini South at Sunset / K. Pu‘uohau-Pummill / HCG 87-GMOS-S

Profile for Gemini Observatory

Issue 30 - June 2005  


Issue 30 - June 2005